Range-azimuth tracking systems and Differential Global Positioning Systems (DGPS) are used to obtain positions of the boat and instruments to within 1 meter. Range-azimuth tracking systems are similar to total stations used for land surveying. The tracking system must be located where the operator can manually track a reflector mounted on the boat. During floods this can be a significant problem. Real-time, kinematic, DGPS allows for rapid collection of velocity and bathymetric data in open areas, but data collection near tree lines and bridges is hampered by loss of adequate satellite coverage caused by blockage of the sky by trees and bridge structure. The latency time between the position and depth measurements can significantly degrade the accuracy of the position measurement and must be adequately treated to obtain the expected accuracy.
The distance from the water-surface to the streambed is commonly measured with an echo sounder. The accuracy of a streambed-elevation measurement is dependent on the echo sounder and the stability of the deployment platform. For hydrographic surveying, it is important that the echo sounder provides signal processing with a digital output that can be recorded by a computer. Two types of signal processing schemes are commonly employed. The most common scheme is threshold detection, which measures the distance based on the time from acoustic release until the reflected acoustic energy exceeds a predetermined threshold. Peak-value detection analyzes all reflected energy and computes the distance associated with the peak of the return signal. Transducers are characterized by their frequency and beamwidth. Higher frequency signals (>200 kHz) provide better resolution but poorer penetration of deep or sediment-laden waters. The acoustic footprint of the echo sounder is a function of the beamwidth of the transducer. A wide beamwidth (>8°) results in a large footprint and less accurate measurements of steep slopes or rapidly changing bottom. The peak detection method is less sensitive to acoustic reflectors in the water column (sediment, fish, debris, etc.) and tends to measure the approximate center of the acoustic footprint rather than the edge of the footprint, effectively reducing the acoustic footprint. The echo sounder currently used for detailed measurements has a 3° beamwidth transducer operating at 200 kHz, digitizes the depth using peak-value detection, and provides both RS-232 compatible digital output and an analog paper chart.
The paper chart records the vertical location of all objects causing an acoustic reflection, while the digital processing produces only a single value. When surveying near objects or in areas with submerged vegetation, false bottom readings are common in the digital data. Therefore, it is important that the paper chart be used to verify the digital data and that any necessary corrections to the digital data are made.
Without vessel motion compensation, the accuracy of the riverbed elevation measurements is a function of the dynamic motion of the boat and slope of the water surface. The instruments designed to measure vessel attitude accurately in dynamic conditions are very expensive and were beyond the financial constraints of this project. Therefore, streambed elevation is determined by assuming the transducer remains vertical and at a fixed distance below the average water surface. Near the bridge, the water surface is assumed constant; in the approach and exit reaches, the water-surface elevation is adjusted for the average water-surface slope in the area estimated from upstream and downstream gages and concurrent water-surface surveys. The estimated accuracy of the riverbed elevation data is about 30 cm.
Deployment of the instruments with a boat is necessary to obtain the spatial coverage required for detailed data sets. The use of manned boats during floods can be hazardous in many situations. Development of an unmanned remote-control boat has eliminated the restrictions of collecting data from a manned boat. The remote-control boat is a 9-foot flat-bottom boat with an 8-horsepower outboard engine, modified to allow control with commonly available recreational radio controls. The boat was modified to provide a wet well in the center for deploying instruments, the transom was reinforced, and additional floatation was added. Data collection from a manned boat is faster and more efficient, but when launching facilities or safety considerations prevent the use of a manned boat, the remote-control boat allows collection of data in areas that were previously inaccessible. All data are radio linked to a field computer located either on shore or on the manned boat so that the position of the instrument and the data collected are recorded simultaneously.
Detailed data on scour at bridges have been collected using this system since 1993. The entire system and field personnel can be deployed within 24 hours to any location in the United States. To date, data have been collected at sites in California, Texas, Missouri, Illinois, Indiana, Iowa, Minnesota, and South Carolina. The remote-control boat has been used successfully during floods in South Carolina, Missouri, Illinois, and Indiana to collect data that would have been dangerous or impossible to collect using a manned boat. The data are currently being used to study scour processes and to evaluate the applicability of various numerical models.